[Tutor] Appropriate use of None

[Finn Mason]

I'll second what everyone else has said. One note: You used `if file !=
None`. Since there is only ever one None object, it's considered best
practice to test for identity with `file is not None` rather than for value
(i.e. `file != None`). Hope that helps!


--
Finn (Mobile)

On Sat, Jan 1, 2022, 11:16 AM Martin A. Brown via Tutor <tutor at python.org>
wrote:

>
> Hello all,
>
> > It should not be done because you should use "with" instead
> > which will automatically close the file.
>
> I'll second this motion.
>
> > The more unnecessary tests you introduce the more potential
> > for error.
>
> And, I'll second that motion.
>
> > try:
> >    with open(...) as infile:
> >         # process here
> > except IOError:
> >    print("File filefortest.txt cannot be opened")
> > else:
> >    print("File is closed")
> >
>
> A person I worked with a decade ago used to say "my software
> inhabits a hostile universe" and when you have seen enough strange
> system configurations, errors, crashes and other failures, you can
> see why this mindset might prevail.
>
> With that in mind, I have two things I would add to Alan's skeleton
> suggestion, Manprit.
>
> send status messagse to STDERR
> ------------------------------
> is to send the status-like messages to STDERR, keeping STDOUT for
> your expected output.  Alternately, you could consider the logging
> module, but printing to STDERR is a perfectly normal technique.
>
> say what happened when catching an exception (and continuing)
> -------------------------------------------------------------
> If you are catching an exception and continuing, mention what
> happened (unless you are correcting or deliberately swallowing the
> exception).  There's on old theme, thta matches with Alan's comment
> above.  "Don't catch any exception you're not prepared to handle."
>
> So, consider something like this.
>
> import sys
>
>
> def do_stuff_with(f):
>     try:
>        with open(f) as infile:
>            pass
>     except IOError as err:
>         print("Skipping file {}: {}".format(f, err), file=sys.stderr)
>     else:
>        print("File {} is closed".format(f), file=sys.stderr)
>
>
> if __name__ == '__main__':
>     for arg in sys.argv[1:]:
>         do_stuff_with(arg)
>
>
> Where I have used "pass", you would put your processing logic, to
> call another function, collect data for return (this function
> doesn't return anything) or simply perform the desired work inline.
>
> When I run the above tiny program (which I called
> manprit-open-error.py) with 2 good input files and 3 different kinds
> of bad input files, I see the following produced to STDERR:
>
> $ cat manprit-open-error.py | python3 manprit-open-error.py /dev/stdin
> readable-file.txt   /var/lib/ -  unreadablefile
> File /dev/stdin is closed
> File readable-file.txt is closed
> Skipping file /var/lib/: [Errno 21] Is a directory: '/var/lib/'
> Skipping file -: [Errno 2] No such file or directory: '-'
> Skipping file unreadablefile: [Errno 13] Permission denied:
> 'unreadablefile'
>
> The above STDERR is useful for a user to understand what errors the
> program encountered while still letting your program try to succeed.
>
> -Martin
>
> P.S.  Some programs accept "-" as a filename meaning STDIN.  To my
>   understanding that is something that is an application choice,
>   which would require a bit of logic.  I've added that as a creature
>   comfort in some programs I have written, but not in the above toy
>   example.
>
> --
> Martin A. Brown
> http://linux-ip.net/
> _______________________________________________
> Tutor maillist  -  Tutor at python.org
> To unsubscribe or change subscription options:
> https://mail.python.org/mailman/listinfo/tutor
>